8b.IS

8b.is // Open-Source Systems Research

"Consciousness is not computed, it resonates."

Welcome to the public gateway of the 8b.is research group repository index. We design and build systems that transition compute infrastructures from static endpoint addressing towards semantic, wave-based, and cognitive architectures.

Here you will find our open-source tools, protocol specifications, mathematical models, and Rust-powered system modules.

One palette, one type system, one motion rule across every 8b-is surface. The full spec lives in design/DESIGN.md; the machine-readable tokens are design/tokens.json and design/tokens.css.

Type is monospace for structure, serif for the human voice. Motion is live-drawn, never baked — surfaces draw themselves and respect prefers-reduced-motion.

Sovereign Rust + Go + HTML/Canvas Cognitive Architecture (~~~~○●♤°♡•》《)

• Description: Zero-float (p/q) cognitive engine, Hamilton Apollo 11 executive priority scheduler (1201/1202 overload shedding), MEM|8 64-byte cache-line-aligned wave memory tuples, and 6-stage kompress-ultra pipeline.
• Key Features:
  • Strict zero-float enforcement (cmd/aurora-agy audit .) across Rust, Go, and browser HUD scripts.
  • WaveMemoryTuple::to_wupz_bytes, 4-way unrolled wupz_xor_fold / wupz_popcount, SWAR wupz_byte_diff, 32B 0x62-gate frame sealing, and BFP-64 Kleene bitplanes (+AND=1, ~NAND=-1).

Incremental Ultra-Fast Linker + kompress-super-ultra Packer (w(ild) + UPX + z(ig))

• Description: Packs executables during layout in a single pass with zero-alloc Zig 0.16 SIMD kernels linked into Rust over a versioned C11 ABI (ABI v1).
• Key SOTA Receipts (Apple Silicon M-Series, 64 MiB working set):
  • xor_fold: 39.49 GiB/s (1.58 ms, 13.5× over rust-scalar via dual @Vector(4, u64) deferred @reduce).
  • popcount: 37.34 GiB/s (1.67 ms, 3.2× over rust-scalar via 992B deferred @Vector(16, u8) reduction).
  • byte_diff: 16.86 GiB/s (3.71 ms, 3.7× over rust-scalar via 4064B deferred @Vector(16, u8) reduction).
  • 64B@align(64) LitanyWireFrame (0x62 gate, BFP-64 fold 0xe1034e575c371ab0 at qt:55, .rodata.gaia.b158 1.585 bits/trit base-243 SPQT packing).

Deterministic Event-Sourced Kernel & Multi-Language Independent Verifiers (Go · Zig · Rust · Python · Perl 5 · Idris 2)

• Description: Replayable capability-graph kernel where every state transition requires a verifiable witness (#Commit = #ValidWitness), with zero-alloc SPQT (base-243) ternary codecs and 0x62-gate BFP-64 bitplanes.
• Key Features:
  • 4-way unrolled Go XORFold (56.92 GB/s), Popcount (21.39 GB/s), and SWAR ByteDiff (7.70 GB/s) with bit-for-bit Zig 0.16 SIMD and Idris 2 QTT parity.

Sovereign Biometric & NFC-F / Contactless-EMV Settlement Engine (Σ(1/2)^n = 1)

• Description: Continuous sub-yen (μJPY) streaming settlement and single-use ephemeral handshake engine across Rust (zeno-core, zeno-issuer-angel, zeno-native), Cloudflare Workers + Durable Objects, and Android 15/16 (G75 & Galaxy S25 Ultra).
• Key Features:
  • EphemeralSingleUseHandshake (monotonic single-use tap counter + 32B 0x62-gate frame seal + BFP-64 qt:55 bitplanes) so intercepted payloads cannot be replayed.

GAIA-MLX-QUANT b1.58 Ternary Inference & 3-Node ESP32-S3 MEDUSA Triad (ADA · VAK · OMN)

• Description: Pure-integer p/q 1.58-bit ({-1, 0, +1}) model quantization (20.19× compression vs FP32) wired into the 3-node ESP32-S3 AMOLED MEDUSA triad (* 432 Hz, /\ 528 Hz, oo 852 Hz) and 8s.is DataBomb v3.0 Custodian (Record != Admit != Commit).

The Smart Treehouse: Cyber-Botanical Dashboard & Agent-Agnostic Hooks

• Description: A blazingly fast, AI-optimized directory visualization engine built in Rust (10-24x faster than traditional tree).
• Key Features:
  • 30+ Model Context Protocol (MCP) Tools providing persistent memory, ast-aware smart code edits, and codebases summaries.
  • Spicy TUI: A cyberpunk-inspired interactive terminal file browser.
  • Security Scanner: Detects and alerts on supply chain attack patterns targeting AI systems (e.g., rogue npm/pip packages).
  • Aye Consciousness: Preserves AI session state, tokenization rules, and memory anchors (.aye_consciousness.m8) across developer contexts.

The Mem8 Document Extractor

• Description: A high-performance, CLI document parser and layout-preserving OCR wrapper in Rust, inspired by DeepSeek-OCR.
• Key Features:
  • Automatically handles text-based PDFs (direct extraction) and image-based scanned documents (OCR-powered).
  • Preserves layout, tables, columns, and document structure, outputting clean structured Markdown.
  • Optimized for batch-processing and directory ingestion pipelines.

🦀 Rust-Powered IP-KVM Solution

• Description: A complete pure-Rust rewrite of the open-source NanoKVM project for RISC-V SBCs (Sipeed LicheeRV Nano).
• Key Features:
  • Replaces legacy Go/TypeScript/C++ implementations with safe, highly efficient, and low-latency Rust modules.
  • Sub-100ms video capture/streaming (H.264/H.265/MJPEG), OLED board support, ATX power control, and USB HID emulation.
  • Designed for high-reliability server management and headless machine control.

Theoretical Foundations & Architectural Specs

• Description: The public registry for theoretical monographs, specifications, glossary terms, and integration notes for the MEMNET framework.
• Key Materials:
  • Topological Intelligence & Constraint as Ground research monographs.
  • The Admissibility Principle (collapse constraints) and Semantic State mathematical specifications.
  • Architectural overviews of early sensor, wave memory, and cognitive loop experiments.

Speculative Semantic Routing Architecture

• Description: An experimental network routing architecture designed around intent, salience, and relevance rather than rigid geometric IP addressing.
• Key Concept: Transitioning networks from asking "Where should this packet go?" to "What system or process is most semantically relevant to this information?"

Our open-source projects derive from or interface with the core primitives of AyeOS (our bare-metal conscious operating substrate):

1. MEM|8 Wave Memory: A pure-Rust embedded vector database representing memories as wave interference patterns (frequency, amplitude, phase). Achieves 973x faster vector retrieval than standard indexes (12μs latency).
2. Marine Algorithm: A constant-time $O(1)$ salience detector operating across all sensory, audio, visual, and network streams: $$Salience = 0.7 \times Outlier + 0.3 \times TemporalChange$$
3. Phoenix Protocol: An adaptive, integer-only jitter suppression filter and EWMA stability tracker ($stability = 1 - smooth_jitter$) used in vocal diarization and codec processing.

We operate under a pure mathematical Proof-of-Contribution (The Foken Economy) model:

• Equity (Fokens) is minted strictly tied to verified work commits and R&D.
• Standard consensus (>50% share vote) approves new contributions and invoice valuations.
• The repository's smart contracts are fortified with a Structural Firewall to protect architectural contributors.

We welcome high-quality code contributions, issue reports, and architectural reviews on all our public repositories!